Organizational stress, job satisfaction and job performance: where do we go from here?

This article reviews and summarizes two decades of empirical literature

concerned with both direct and moderating variable-based analyses

of the relationship of organizational stress with job satisfaction

and job performance. Moderating influences of various constructs

operationalized at the individual, group and organizational level of

analysis are classified and then reviewed systematically. An evaluative

summary, of this research suggests that although there have been significant

improvements in the analytical methods employed to investigate

such phenomena, much of this research still does not consider the role

of reciprocal relationships that evolve over time. We provide four

guidelines for improving the quality of both theoretical rigor and

methodological robustness in this important area of organizational

inquiry.

Despite recent advances in research on organizational stress, there is considerable disagreement on basic theoretical and methodological issues. As a domain of scientific inquiry, this area has been of central importance to the field of organizational sciences. The belief that stresses experienced by individuals can affect important organizational outcomes is shared by numerous researchers. Ivancevich and Matteson (1980) provided estimates of stress-related costs to the U.S. economy that if reported in present day dollars would be much over 10% of the GNP. Although most of such costs are due to the health-related injuries and mental stresses encountered in organizational contexts, there is considerable loss due to effects of stress on important organizationally valued outcomes, such as job satisfaction and job performance. In this review, we examine the relationship of organizational stress with these two important outcomes. First, we review and analyze the empirical studies conducted in the past two decades that relate organizational stress with two outcomes. Second, we provide an evaluative summary of much of this research. Third, we suggest four guidelines that are likely to aid in improving both theoretical rigor and methodological robustness in future research designs. It is hoped that the quality of future research findings will be greatly enhanced if researchers pay some attention to these four concise guidelines that we have developed at the end of our review.

Because reviews of the history and scope of research on organizational stress are already available elsewhere, (see Beehr & Bhagat, 1985; Ganster & Schaubroeck, 1991 among others), it is not necessary to engage in such an endeavor. Instead, we will begin our review of the relationships of organizational stress with valued organizational outcomes in a systematic fashion. First, we review the relationship between organizational stress and job satisfaction focusing on both its direct or main effects and then on moderating influences operationalized at the individual, group, and organizational level of analyses. Second, we review the literature concerned with the relationship between organizational stress and job performance. In this section, just like in the previous one, we analyze the significance of research efforts that use moderating influences of various types. However, because there have been relatively fewer investigations dealing with such latter types of concerns, our focus has been mostly on direct relationships between stress and job performance.

Relationship between Organizational Stress and Job Satisfaction

Most of the research on organizational stress has focused on its relationship with job satisfaction. Much of this research has been correlational studies that have used role ambiguity and role conflict to operationalize stress. These studies generally indicate that job stress and satisfaction are inversely related (e.g., Hollon Chesser, 1976; Miles, 1976; Miles & Petty, 1975). Because the relationships between role conflict and ambiguity, and organizational outcomes have been meta-analyzed (Fisher & Gitelson, 1983; Jackson & Schuler, 1985) reviewed (Van Sell, Brief, & Schuler, 1981), and critiqued (King & King, 1990) elsewhere, our analysis does not include these previously examined areas. Instead we focus on newer methods of data analysis and other operationalizations of job stress and research completed since these meta-analyses.

In addition to these correlational studies, more sophisticated techniques, such as Lisrel and path analysis, have been used to examine the stress-satisfaction relationship. For instance, Kemery, Mossholder, and Bedeian (1987) employed Lisrel to test three models, (e.g., Beehr & Newman, 1978; Locke, 1976; Schuler, 1982) that postulate causal relationships among role ambiguity, role conflict, and organizationally valued outcomes such as job satisfaction, physical symptoms, and turnover intentions. Using 370 employees, (e.g., faculty, administrators, staff), from a large southeastern university, Kemery and associates found that role conflict and ambiguity exert a direct influence on job satisfaction and physical symptoms, which in turn influence turnover intentions.

Similar findings of the indirect effect of stress on turnover intentions through job satisfaction have been reported by Hendtix, Ovalle, and Troxier (1985) and Kemery, Bedian, Mossholder, and Touliatos (1985). Hendrix and associates used a stress assessment package developed by them to measure organizational stress, job satisfaction and turnover intentions of employees working for the Department of Defense (n=341) and a civilian hospital (n=29). They did not find a direct, significant relationship between organizational stress and turnover intentions. However, results of a path analysis indicated that job satisfaction was affected by factors such as involvement in decision making, skill variety, and whether work was subject to the whims of supervisors. In turn, job satisfaction was strongly linked to the intention to quit. Kemery and associate (1985) used three samples of accountants, (public n=275, government n=254, industrial n=459), and a sample of hospital employees (n=66) derived from Jackson's 1983 study, to replicate the Bedeian and Armenakis (1981) model of the relationship between role ambiguity and conflict, as well as job tension, satisfaction, and intention to leave. Using Lisrel, they found that data from these three samples of accountants supported the Bedeian and Armenakis model(1). Stress exerted an indirect influence on turnover intentions through job satisfaction. However, unlike the results of Kemery et al. (1987) and Hendrix et al. (1985), stress also exerted a direct influence not only on job-related tension and job satisfaction, but on the propensity to leave the organization. The diversity of job types, (e.g., university, defense department, hospital employees), as well as the differences in experienced stress levels and in the measures employed could explain these conflicting findings.

Relative effects of different sources of stress on job satisfaction have also been analyzed. Drory and Shamir (1988) examined the effects of intraorganizational factors, (e.g., role conflict, role ambiguity, management support), extraorganizational factors, (e.g., community support, family-role conflict), and task characteristics on the job satisfaction and burnout of 266 Israeli prison guards. They found that extraorganizational factors, especially community support, made the greatest contribution to explained variance (12%) in job satisfaction. Task characteristics accounted for 4.35% and organizational variables accounted for 3.4% of the explained variance. These results suggest that extraorganizational types of stresses are as important as intraorganizational sources in determining an individual's levels of job satisfaction in Israel. These results also emphasize the impact of the non-work factors on work outcomes in the Israeli context. However, in explaining job burnout, intraorganizational factors accounted for 9% and extraorganizational factors accounted for 5% of the variance. Task characteristics did not add significantly to the net explained variance in job burnout. Although both intra- and extraorganizational factors made unique contributions to the explained variance in burnout, internal factors accounted for a greater proportion. These results seem to suggest that though management support may be an important factor in preventing burnout, it is probably of lesser importance in preventing job dissatisfaction.

More recently, researchers have begun to examine variables that might moderate the relationship between organizational stress and job satisfaction. In Table 1 we list the complete set of possible moderating variables contained in the articles reviewed. As depicted in this table, most of these moderating influences have been examined at the individual level of analysis and have focused primarily on job satisfaction. Although we list other outcomes variables in Table I as well, the analysis that follows focuses on job satisfaction and job performance respectively.

Moderating Influences: Individual Level of Analyses

Numerous individual level variables have been examined as potential moderators of the relationship between organizational stress and job satisfaction. For example, Bhagat and Allie (1989) examined the moderating effect of sense of competence on the stress-satisfaction relationship of 276 elementary school teachers. They found that when organizational stress was high, individuals with a high sense of competence reported greater satisfaction with work and co-workers and reduced feelings of depersonalization, compared to those with lower sense of competence. When experienced stress was low, highly competent individuals were less satisfied with co-workers than were individuals with a low sense of competence. One's sense of competence also moderated the effects of personal life stress on organizational outcomes. Under conditions of high life stress, highly competent individuals reported greater satisfaction with work, co-workers and supervision, less emotional exhaustion, and less feelings of depersonalization than did individuals who perceived themselves to be less competent.

In addition to sense of competence, the moderating effect of perceived control on the stress-satisfaction relationship has been examined in the following studies. For example, Tetrick and LaRocco (1987) employed a sample of 206 physicians, dentists, and nurses from a naval hospital to investigate this issue. They examined the role of the ability to understand why and how organizational events happen, to predict the frequency, timing and duration of such events, and to control important outcomes by influencing events and significant others in the work environment. They found that such perceived control could indeed moderate the stress-satisfaction relationship. However, the ability to predict events did not moderate the stress-satisfaction relationship.

Conflicting results have been reported on the moderating effects of …